ConferenceRoland

I would like to place this lecture in the context of an aphorism written by Mallarmé:

"To wit, social relationships and their momentary measure, be it constricted or expanded in order to rule, is a fiction, which depends on the Belles-Lettres ."

Mallarmé points here to a significant fact: that the political order ultimately depends, for its founding, upon a fiction. This fiction may be of the order of an historical document or, even, a poetic text. This should not come as a surprise: after all, any newness, in the human domain, depends on a linguistic act. These acts can be expressed in charts, edicts, etc., but also in poetry. We can surmise that, because of the power of its persuasive rhetoric, a poetic text has an advantage over the dryness of an historical document.

Of course, a lengthy explanation about my conception of fiction is in order here; but I will simply refer you to Jeremy Bentham and his remarkable Theory of Fictions , in which he establishes that fiction is capable of creating truth, in whatever field it is used (philosophy and politics, but also of science, which for him are the primary domains of experience for the fictitious).

I have no valid reason to go against the long medievalist tradition that has held the Song of Roland as a founding text since its discovery, in 1832, at the Bodleian library. Neither do I have a qualm about what is founded, namely: France itself. But then, questions arise:

Of course, it would be impossible to treat these "big" questions exhaustively in a short (I hope) conference, but I will try to give you a set of concise, schematic answers.

The Song of Roland indeed creates, from an international, polylinguistic feudalism, France as a monolinguistic Nation-State. It is the first text to promote the name "France" to writing in the vernacular old French: the founding of the Nation has therefore an essential relationship to writing. It is through writing that the Song enacts the substitution of the feudal pyramid by the monarchy with a unique ruler and a mass of subjects.

The change in the political structure (in the form of the Law) has psychological consequences. To frame them, I will refer to the Freudian vision of society and politics, as expounded in Civilization and its Discontents or Group Psychology and the Analysis of the Ego . The birth of the group, for Freud, begs a question that is akin to the one of the chicken and the egg: which came first? The Freudian answer, as far as society is concerned, is the following: political institutions are a projection of the superego: simultaneously, society, through language and parental authority, determines the historical form of this very superego. We are here faced by a double causality, where the superego produces what produces it. This explains why, in parenthesis, political change is so hard to achieve and at the same time full of unintended consequences: since our unconscious collaborates with the social order, no wonder political institutions fiercely resist our conscious attempts at changing them.

Let us go back to the schema: in the feudal pyramid, the ruler is dependent on the vassals, he is not the absolute father of the primal herd, but a "first among equals," as proven by the way Charlemagne, at the beginning of the text, always defers to his vassals' council. On the other hand, in the end Charlemagne has constituted himself as the Absolute (or dead) father: everybody, including the noblest sons of France, depends on him as a subject.

But how is this new political and psychological order created? This is where the notion of homonymization comes into play (forgive me this barbaric sounding designation). Newness, in human history, emerges by the power of a new name: a homonym. Let me give you an example borrowed from the hard sciences: the word "orb", in the Ptolemaic cosmology, designates a perfect circle. The very same word, in Copernicus, points to a very different concept: the ellipsis of the planets. Even if the words are the same, they do not designate the same thing: they are homonyms.

Applied to the Song of Roland, this bit of epistemology can be stated thus: even if the words describing power and subjection are the same (emperor, king, vassal, subject), their content radically differs in the beginning vs. the end of the text. In fact, the text has brought about a radical alteration of their meaning; it has produced their homonymization.

Hence, now, and not without some irony, the notion of a virtuous circle. The text moves along the circle of names. By this very revolution, it does not repeat itself: quite the contrary, the circuitous path of the repetition of the old names leads to the creation of new names.

This may sound very abstract, so let me get immediately into a concrete reading of the text; let me begin by the characters are regularly (and, in my view, mistakenly) viewed as the "Others" of the text, the Saracens.

Saracens in the Song of Roland do not refer to any historical reality. They are polytheistic; they adore idols and have therefore nothing to do with the monotheistic, iconoclastic Muslims whom the Crusaders confront at the end of the XI th century. In fact they have nothing to do with any enemy history could have supplied to the budding French monarchy of the time.

What is then the textual genealogy of these fictitious pagans? I will not rely on the easy argument of the author's ignorance - always a costly mistake among medievalists or others. In fact, the Saracens of the epic are the anachronistic resurgence of a topo known since Saint Paul and elaborated by Tertullian and Saint Augustine; they are emblems of the idolatry of antique paganism.

Let me remind you of the most important points of the Church Fathers' critique leveled against the cult of idols. Idols are deified men, non-transcendental deities. Their simulacra (statues, images, fictional representation on the histrionic stage) are fabricated by man. For the Apostolic Fathers, the pagans render a cult to themselves, instead of adoring a transcendent deity. In this polemical anthropologization, idols are but a circular, narcissistic fiction representing only the immanence of human desire. This aspect of human fabrication is indexed by the Song when Baligant, chief of the Saracens' army, promises to cover the gods' images with gold (l. 3490 sqq).

For the Church Fathers, the idols also constitute the space where man encounters the devil. Let us quote Tertullian:

"Will we find something more privileged to encounter the Devils, his pomp and his angels than idolatry? It is through idolatry that any bad and vile spirit, I daresay, secures our obeisance."( De Spectaculis , IV, 2)

Idolatry, then, is also a demonology. This is underlined in many instances by the Song of Roland . Examples of Satanism abound in the Saracen camp (cf. L. 980, 1268, 1390, etc.).

But this fictitious paganism is not limited to a religious content. It is also a certain mode of fictional writing, in the sense that, for the Church Fathers, Satan is before all a producer of simulacra, the artisan of a theologia theatrica , as Saint Augustine has it (De Civ. Dei , VI, vii). Lucifer is before all an artist, an interpolating and rival angel/writer who rewrites and ruins the book of divine creation ( Tertullian , DS, II, 7). Satan, then, can be conceived as the inverted and secret model of an entire section of medieval literature, and the pagans in the Song of Roland metaphorize the writing of a satanic fiction. On another level, quite literally, pagans generously revive the topos of the blood writing, by supplying their blood/ink to the Christians; the latter dip in this blood their sword/pen, as do their enemies, all this to write an epos. In particular, Ethiopians are blacker than ink.

This demoniacal writing is grounded in a process shown explicitly in the first lines of the text, when the Saracen Blancandrin advises his lord Marsile. I quote:

To secure the survival of the Saracen social order, and his eminent position inside it, Blancandrin is ready to sacrifice his own son. Let us now superpose upon his words a quotation from the Gospel:

At the heart of Saracen society reigns the victimization process so aptly described by Rene Girard. The scapegoat, in Jerusalem as well as in Saragoza, seals by his expulsion or assassination the restoration of the threatened social order - or, in the case of the Song of Roland , the founding of a new social order.

I posit here a hypothesis: the so-called "pagan" ethical and scriptural model that fabricates idols and that grounds itself in the scapegoat's blood is the one that informs on the most general level the Song of Roland in its entirety.

I know, of course, that this hypothesis goes against the tendency of one and a half centuries of critical tradition that never ceased to stress the radical antithesis between the pagan and the Christian worlds, assimilated to Evil and Good. However, some inspired precursors, like Yves Bonnefoy and more recently Peter Haidu, in his magnificent book The Subject of Violence , have begun to break with this uncritical commonplace. The question to be asked to this interpretative tradition is the following: Isn't the difference between pagans and Christians a diacritical one, that is a fictional one, constituted by a textual play to ends we still have to identify? Isn't it in the interest of the Song to make this diacritical difference appear as an essential one?

The first element of an answer resides in the fact that the Song of Roland itself is under the empire of the simulacrum it denounces. In particular, by creating ex nihilo a Saracen paganism that has nothing to do with historical reality, the text is engaged in the process of demoniacal writing, and cannot be distinguished from the simulating idolatry it imputes to the "other" side. By the way, this process of accusing the adversary of the crime one commits, is inventoried in classical rhetoric under the name metastasis (a cancerous process).

But the workings of the simulacrum are not, by any means, restricted to the creation of the Saracens: Roland, Charlemagne, Olivier, Turpin, these XI th -century feudal lords have of course very little to do with the real Carolingians. Also, Peter Haidu has shown that Charlemagne creates with his breath only a whole army: one-hundred-thousand French men, whereas all the French have been massacred at Roncevaux. Those new warriors are his French indeed, as the text underlines (3405). The Song of Roland transforms the emperor into a usurper of the creating Word of Genesis, a word that proffers and performatively creates something where there was nothing.

To penetrate further in this act of creative fiction, let me examine briefly a singular textual inconsistency - and, again, I follow Bentham: fictions are not lies or illusions, or fantasies: they can reveal a kernel of truth.

As the text indicates, all the combatants at Roncevaux are dead. How is it then that the text can write?

This is indeed an amazing miracle! From where does this surviving feudal lord come, the phantom that exists without really existing, this pure fiction whose vacuity the song in fact stresses? However, the destiny of meaning seems to depend on the charter that authenticates the text itself: without it, we are told, we will never understand anything. In fact, by the visual testimony of the noble Gilles (a word that means "deception" in Old French), this pure "fiction that avers the truth", the Geste authenticates itself in a circular manner, by referring at the same time to its own fictionality.

On a totally different level, the Franks, at the beginning of the text and the Saracens as a whole share the same values, an identical ethics and ideology: the ones of the European feudal system. In that respect, even Ganelon, the "traitor," and his family show no difference from, for example, Roland and Olivier. As Peter Haidu has shown, the feudal value system in the Song of Roland is submitted to a radical reevaluation that shows its catastrophic failure. First (because of "honor"), it leads the twelve French peers to their deaths at Roncevaux. Also, the reciprocal obligation of auxilium (help) between the vassal and the suzerain is never honored by the lords of the text; For example, the pagan lord Baligant has left his vassal Marsile to struggle against Charlemagne for seven years, the Saracen gods cannot help the pagans' defeat, and, particularly, Charlemagne abandons Roland to his death. The feudal warrantors are a complete failure, which is recognized by Charlemagne himself when he speaks to Roland's corpse:

The text plays here on the ambiguity of "por" in Old French, which can indicate either the cause or the destination (an ambiguity present in English too, in the preposition "for").

Moreover, the differential mark constituted by religion is undermined in the text: The Christianity of the Song has very little to do with history: for example, the archbishop Turpin is but a warrior disguised as a priest. In brief, Christianity in the text is nothing else than a theologia theatrica . Therefore, one cannot ground an essential difference on it; the textual functioning of Christianity only points to a play of diacritical oppositions built by the epos itself.

As a matter of fact, the pagans of the text speak and act like good vassals, " a lei de bon vassal" (887): that is, they are the mirror images of the feudal virtues exemplified by the opposite camp. But the similarity goes even further. We have seen that Saracen society is grounded on the scapegoat mechanism. But isn't it the same for Charlemagne? In fact, we must consider Roland and the twelve peers as scapegoats: They have been knowingly sent to their death by the suzerain; when Ganelon designates Roland for the ambush of the rearguard, (which, according to feudal law, he has scrupulously and publicly warned about), Charlemagne does nothing, or rather he shows that he knows that Roland is sent to his death. Also, when he offers the half of his army to his nephew, he knows that the latter, because his concept of honor, will decline the offer.

Aren't Marsile and his vassals, but also Baligant and his one million and five hundred thousand men scapegoats too? And what about Ganelon, who is exonerated of the crime of treason by the emperor's council, but condemned to be quartered in virtue of a new law at the end of the text, created ex nihilo by Charlemagne?

All those innocent victims, on one side or the other, share a common characteristic: they all belong to the class of the feudal lords. In consequence, at the end of the Song , Charlemagne has succeeded in exterminating the feudal class in its entirety, be it Christian or pagan, in order to cement a new society in the scapegoats' blood, a society that is none other than the monarchy of the nation-state.

In that sense, the moment that seems to insure the triumph of Christianity over idolatry - when Charlemagne shatters the idols in Saragossa - is in fact the moment when a new, secret idol is erected: Charlemagne, as the idol of the nation-state.

This is demonstrated by the circular structure of the text: with Roland's, the Saracens' and Ganelon's holocausts, Charlemagne occupies the place of the persecutor in the scapegoat mechanism, this space defined at the beginning by the pagan Blancandrin:

As a matter of fact, Charlemagne follows in all his acts the model of Saracen writing, this paradigm grounded on simulacrum and persecution. He is the ultimate idol hidden by a theologia theatrica that disguises itself by the rhetoric of a dubious Christianity. The text, therefore, ends where it had begun, its end engendering its beginning and vice versa. The commutation of values from one side to the other and the rhetorical inversion imply that the discourse of the pagan other is identical to the discourse of the Christian sameness; reciprocally, the pseudo-Christian discourse is only a copy of the Pagan one.

Are we then confronted by the vicious circle of an infinite repetition? Absolutely not: in going through the circle of the simulacrum and the scapegoat for a second time, Charlemagne has metamorphosed many significations: Strategic signifiers like "France", "vassal", "emperor", "king" have been submitted to a tremendous homonymization. For example, the feudal lord of the beginning, who must submit to the advice ( consilium ) of the peers' council (and has always done so, until the end of the text), suddenly becomes an absolute monarch who invalidates the advice of his own council and promulgates a new law. When Charlemagne refuses to reconcile with Ganelon and inculpates him through an ordeal, the text operates what is nothing else than a coup d'état.

If the names, and the reality attached to them, change, therefore we are in the space of a virtuous circle. This virtuous circle can be justifiably called a revolution, from revolvere : the turning around of names. The Song of Roland hereby indicates that the empty case of the ruler doesn't move and never ceases to exist. What changes are the successive usurpers of the space of power, producing historical transformation?

Now some conclusions:

First, instead of passively reflecting a historical reality, the Song of Roland prophetically anticipates a coming reality: the one of the monarchical nation-state, even before the charters register the change - it is only in 1181 that the charters name the king rex Franciae instead of rex Francorum . The text is the foreboding space of an active metamorphosis between two social orders.

Second, far from being a monosemic glorification - as criticism believed for a long time, the epos constitutes in fact a radical critique of power mechanisms, of ideological use of religion, of politics in general. The text subversively shows that political meanings, instead of being univocal, are always ambiguous and that this ambiguity has to be resolved by the violence of an arbitrary and, at the same time, necessary decision.

Third, the Song of Roland knows that the monarchy of divine right, which it gives birth to, obeys to a structure of simulacrum, that the monarch draws his power, not from a divine unction, but from the skillful manipulations of images weighted by truth, power and history. And here lies the answer to my third question about the relevance of the text: it shows, with a lucidity without equal, the underside of the cards that are still dealt nowadays, be it in politics, marketing, academia, etc.

And finally: if we are to evaluate the historical change formalized by the Song in its consequences, we can distinguish great benefits as well as grievous harms. Among the benefits, we can list the gradual disappearance of feudal violence as a rule of law, its marginalization that makes this violence efficient, in our society, only in reduced sects, like the Mafia or drug gangs. On the side of damages, everybody will have in mind a litany of woes brought by the emergence of the Nation state; war on a planetary scale, the extension of the power of the State over individuals, and so on.

Homonymization and the "virtuous" circle in the "Chanson de Roland"

Now some conclusions: